Python API call to Retrieve Order Book Data from A7 API

This article is designed to show users how to use A7 Python API to extract data (e.g. top of the book) for a given futures contract. We cover a simple call to download data using algos functionality in A7. Addtionally this article provides the backbone of further uses cases.

The following segment of code sets up the API parameters and imports packages.

import pandas as pd
import requests
import numpy as np
from pprint import pprint

proxies = {
"http":  "", #Enter http Proxy if needed",
"https": ""  #Enter https Proxy if needed",
}
owner = "login" # Enter your login name here
 
API_TOKEN="Bearer "+"Enter API Token" 

Below is the main function call to retrieve top of the order book for FGBL futures contract (Security ID 4611674) given a date. Note that a market segment ID (688) needs to be provided in the call as well. This data can be obtained from the A7 UI or from RDI (please refer to RDI usecase).

url = 'https://a7.deutsche-boerse.com/api/v1/algo/{}'.format(owner)
r = requests.get(url = url,  headers={'Authorization': API_TOKEN}, proxies = proxies)
pprint(r.json())

{'algos': ['DBAG',
           'XEUR_Depth',
           'XEUR_LAMQ',
           'XEUR_RPD',
           'XEUR_SLM',
           'XEUR_Spread',
           'auction',
           'depth',
           'instrument_state',
           'price_level',
           'spread',
           'summary',
           'top_level',
           'traded_volume',
           'DBAG',
           'XEUR_Depth',
           'XEUR_LAMQ',
           'XEUR_RPD',
           'XEUR_SLM',
           'XEUR_Spread',
           'auction',
           'depth',
           'instrument_state',
           'price_level',
           'spread',
           'summary',
           'top_level',
           'traded_volume'],
 'success': True}

url = 'https://a7.deutsche-boerse.com/api/v1/algo/{}/top_level/'.format(owner)
url = url+'run?marketId=XEUR&date=20200804&marketSegmentId=688&securityId=4611674'
r = requests.get(url = url,  headers={'Authorization': API_TOKEN}, proxies = proxies)
res = r.json()

The request returns a JSON data format. This can be parsed into a data structure. For example below we expand the JSON format into a DataFrame for the bid side.

df_b = pd.DataFrame()
df_b = df_b.assign(ts=res[0]['series'][0]['content']['ts'])
df_b = df_b.assign(pv=res[0]['series'][0]['content']['price'])
df_b['pv']=df_b['pv'].astype(float)/1e3
df_b = df_b.dropna()
df_b['ts'] = df_b['ts'].astype(np.int64)
df_b['ts'] = pd.to_datetime(df_b['ts'])
print(df_b.head())

                             ts      pv
0 2020-08-04 00:15:13.378085871  177.31
1 2020-08-04 00:15:13.378958683  177.32
2 2020-08-04 00:15:13.474408899  177.33
3 2020-08-04 00:15:13.475364501  177.32
4 2020-08-04 00:15:13.484721498  177.33

The same strategy can be used to extract for the offer side.

df_a = pd.DataFrame()
df_a = df_a.assign(ts=res[0]['series'][1]['content']['ts'])
df_a = df_a.assign(pv=res[0]['series'][1]['content']['price'])
df_a['pv']=df_a['pv'].astype(float)/1e3
df_a = df_a.dropna()
df_a['ts'] = df_a['ts'].astype(np.int64)
df_a['ts'] = pd.to_datetime(df_a['ts'])

Once the data has been extracted and formatted as above it can be visualised.

import matplotlib.pyplot as plt
fig = plt.figure(figsize=(20,15))
ax1 = fig.add_subplot(111)
ax1.step(df_b['ts'], df_b['pv'], alpha=0.9, linewidth=1.0, label='Bid', color='b')
ax1.step(df_a['ts'], df_a['pv'], alpha=0.5, linewidth=1.0, label='Ask', color='r')
ax1.legend(loc="best")
ax1.set_xlabel('Time')
ax1.set_ylabel('Price')
plt.show()